When displaying an image represented by an input analog video signal by the use of a cathode ray tube (hereinafter referred to as a CRT) or the like having a large number of scanning lines in an enlarged or high-definition mode, generally, the input video signal is converted by an A-D converter into a digital signal, the digital data corresponding to one frame is stored in a first frame memory, the data held in this frame memory is converted by an enlarge interpolator into an enlarged image data signal and written in a second frame memory, and then the image data held in this second frame memory is successively read out and converted by a D-A converter into an analog signal, this analog signal output being used in displaying the original image in the enlarged or high-definition mode. When converting the original image data made up of 512.times.512 pixels into the enlarged image data made up of 1024.times.1024 pixels, the enlarge interpolator is designed to interpolate, between adjacent original pixels, a new pixel having such an intensity level as to linearly interpolate the intensity levels of these adjacent original pixels.
However, if such an interpolation processing is adopted in handling the image data on which a so-called two-state image is overlaid, for example, a variable-density image on which another image of character, graphic symbol, cursor, etc. is displayed in superposed form, the interpolation processing is performed in the surroundings of each overlay pixel, or between the overlay pixel having a high intensity level and each pixel of the variable-density image generally having a lower intensity level than the former; as a result, a new pixel having an intermediate intensity level between them is interpolated. Therefore, if the thus interpolated enlarged image data is displayed, the surroundings of the overlay image grow dim, spoiling the sharpness of a display of character, graphic symbol, cursor, etc.